Decision-making In Different Cultures: Essays In Experimental Economics

Doctoral Thesis School of Social Sciences UNIVERSITY OF TRENTO

Decision-making In Different Cultures: Essays In Experimental

Economics

Bassim Abdullah H. Allaheeb

A thesis submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Economics and Management

i

Abstract

eople interact with each other rationally and irrationally. Standard economic theories assume that people act rationally, yet, behavioral economic theories indicate that sometimes people act irrationally because of the specific situation, their character, and many other factors.

Culture plays a role in shaping people’s preferences, values, personalities, norms, and beliefs. That being the case, it is important to shed some light on the effect of different types of cultures on people’s decision-making behavior, in order to better understand human nature with regard to economic decisions.

The ultimate goal of this thesis is to define how people from various cultures behave differently while making economic decisions, and to provide a better understanding for the motives behind people’s preferences towards decision-making.

The first and second chapters of this thesis consider studies in experimental economics about the effect of introducing effort to the ultimatum game. The first chapter provides evidence that people’s decisions are driven by either hierarchy or property rights by comparing the results of three different cultures. The second chapter shows the impact of proposers’ effort in the ultimatum game, in which, culturally driven phenomenon impact on the proposers’ behavior. The third chapter investigates the impact of the Islam religion on experimental studies with regard to decision making and it shows that religious beliefs play an important role in shaping people’s preferences.

Keywords: behavioral economics; experimental economics; ultimatum game; fairness; effort; culture; hierarchy; property rights; Islam.

ii

Dedication

To my mom and dad, who have constantly given me support, unconditional love, and encouragement.

To my wife, who has sacrificed her time, given me her endless support and kindness, and accepted to join me in our life overseas in order to accomplish my goals.

To my siblings: Bander, Lamya, Amal, and Lina, who have always encouraged me on my Ph.D. journey.

To Abdullah, my beautiful son, who puts a smile on my face every morning.

To my true friends, Bandar Alashab, Prof. Abdullah Alberaidi, Dr. Ahmad Alkubair, Abdulaziz Altowaijri, Dr. Adnan Alobiad, and Saleh Alkhidhr, who have kept urging me to complete my thesis.

To my employer, Majmaah University, who gave me the opportunity and supported me with the scholarship that covered my studies and my living costs during all the years that I have spent in Italy.

iii

ACKNOWLEDGEMENTS

This amazing and unique Ph.D. journey that I have gone through would not have happened without the endless support, advice, and enthusiasm from my advisors Prof. Luigi Mittone and Prof. Matteo Ploner, who will both remain in my thoughts and will be remembered forever in my memory. Their great effort, guidance, and encouragement have changed my life, sharpened my critical thinking, and enhanced my research abilities. I am grateful for the knowledge they have transferred to me, and for answering my repeated questions without any hesitation or delay. Both of them have put their valuable knowledge, experience, and wisdom at my disposal at any time.

My sincere thanks to my employer, Majmaah University, who gave me the opportunity and supported me financially to accomplish my Ph.D. abroad in a highly respected university, the University of Trento.

I extend my thanks to the referees, Prof. Luca Corazzini, and Prof. Dirar E. E. Ahmed, who both spent their priceless time in reviewing and providing me with their valuable feedback about my thesis. From their notes, I have learnt and enhanced my knowledge regarding academic ways of writing and reporting. I will owe them for my life regarding their inputs that reshaped my research ability to be in a better place.

I also extend my gratitude to my course mates, who took part in their own Ph.D. journey alongside me, Ali Seyhun Saral, and Tatiana Balmus, who never stopped supporting and

encouraging me to finalize my thesis. My thanks also to the genius and talented programmer Marco Tecilla, who helped me in developing the necessary codes to launch my experiments successfully.

iv

Finally, this Ph.D. thesis would not have been completed without the welcome I received from the University of Xiamen in China, especially professor Yun Wang, who allowed me to use the FEEL lab to conduct my experiments. I am also grateful to the Dean of the College of Ghat at Majmaah University, professor Omar Alomar, and his team members, who helped me to run my experiments in their labs.

v

Table of Contents

1.9.1. The UG with responders’ claims and efforts (UGT) experiment: ...20

1.9.2. The UG with claims but without effort (UGC) experiment: ...23

1.10. Social Value Orientation (SVO) : ... 27

1.11. Further Investigation ... 29

1.12. Recommendations for future studies ... 29

1.13. Limitations ... 29 1.14. Conclusions ... 30 Chapter 2 ... 33 Abstract ... 34 2.1. Introduction ... 35 2.2. Existing literature ... 37 2.3. Target Population: ... 39 2.4. Research Questions ... 40 2.5. Hypotheses ... 40 2.6. Data Analysis ... 40 2.7. Experimental Design ... 41 2.8. Laboratory Environment ... 42 2.9. Offer Making ... 44

vi

2.10. Results ... 45

2.10.A. The UG with a choice to exert effort or not (UGT) : ...48

2.10.B. The UG with exerting effort as a mandatory (UGC) ...54

2.11. Conclusion ... 59

Chapter 3 ... 61

Abstract ... 62

3.1. Introduction ... 63

3.1.1 The Religion of Islam ...64

3.1.2. Prohibited Transactions in Islam ...66

3.2. Literature Review ... 68

3.3. Data Analysis ... 69

3.3.1 The whole sample report ...69

3.3.2. Gender ...69

3.3.3. Age Grouping ...70

3.3.4. Education: Degree and Non-degree ...70

3.3.5. Employment: Employed and Non-employed ...71

3.3.6. Nationalities: Saudi vs Others ...71

3.3.7. Religion ...72

3.3.8. Education and Gender ...72

3.3.9. Grouping of Education, Gender, and Age: ...73

3.4. Research questions and hypotheses ... 73

Hypotheses: ...74

3.4.A. Willingness to take part in experiments that are against religious teachings ...74

3.4.B. Willingness to take part in experiments that are not against religious teachings ...82

3.4.C. Participants willingness to accept cash prizes as a reward ...88

3.4.D. Participants willingness to be part of a game-type gambling?...90

3.4.E. Participants willingness to be part of a gambling-type game that doesn’t use real money ...92

3.4.F. Uncertainty and participant’s willingness to participate in experiments: ...94

3.5. How do participants suggest they play if they were taking part in the Ultimatum game? ...96

3.6. Limitations ... 99 3.7. Conclusion ... 99 CONCLUSIONS ... 101 Appendix A ... 104 Appendix B: ... 105 Appendix C ... 114 References ... 122

1

Introduction

Humans are born in different cultures and are exposed to different types of norms, education levels, philosophies, experiences, and beliefs. Some are driven by individualistic preferences and others are driven by collectivism. Some are shaped by hierarchy and others are property rights’ seekers. Examining cross-cultural differences especially in the current

technological era, in the field of experimental economics, is an essential step towards discovering the nature and real motives of the human decision-making processes to better understand why and how a decision is made. The standard economics theories assume that all humans make their economic decisions rationally, and share the same approach of maximizing self-interest. Yet, behavioral economic studies show that humans are not always rational, in fact, they are acting according to a bounded rationality, for instance, Herbert A. Simon (1955, 1979) argued that decisions are made to gain self-satisfaction rather than maximizing self-interest. Furthermore, some decisions are made to punish or compete with others even if such decisions can lead to losing gains. Since beginning to investigate human motives towards certain types of decisions, economists have been studying the impact of real-life factors that may or may not affect the process of decision-making. Age, gender, education, and other demographic variables have been used to identify human preferences more precisely and to put forward theories about economic drivers.

In this thesis, the main goal is to shed some light on the difference in decision-making preferences between three different cultures, which are Italian, Chinese, and Saudi Arabian, all of them being members of the G20. In order to measure the differences, the effort factor was employed, as a tool, which been used in the literature of experimental economics as a proxy to understand the reasons behind choosing a specific option, and to mimic real life situations. The

2

effort component is focused on in the two first chapters, by introducing it in the Ultimatum Game (UG), which is one of the most famous experimental games in the field. Results of these experiments varied significantly based on how people perceive endowments and the country they come from. Furthermore, several variants were identified as influencing their decisions.

This research aims to provide behavioral insights for researchers, above all, enabling them to take into consideration cultural differences when they design their experimental studies, especially, when experiments conducted in a society that may be driven by a hierarchy structure like Saudi Arabia and China, or in a society where people seek to extend their property rights such as Italy. The first chapter examines the impact of introducing effort to the responders in the UG. By adapting this method, this study was able to detect how participants perceive the

allocated endowment and to what extent the endowment was perceived by people from all these different cultures. Effort was also used as a proxy in the second chapter to analyze subjects’ behavior when effort was exerted by proposers and to what extent those proposers’ decisions differed culturally. The third chapter of this thesis discusses the impact of belief in Islam on people’s preferences in Saudi Arabia, regarding decision-making. This chapter was written as a response to the first two chapters, where unprecedented results were seen in the effort component in identifying why and how participants from a society like Saudi Arabia, which is a well-known as a conservative Islamic country, would act towards experimental studies, i.e. experimental economics studies. An online questionnaire was used to analyze peoples’ preferences and willingness to participate in experimental economics trials. The three chapters all show how important it is for a researcher to recognize the impact of cultural differences in reaction to experimental economic studies and give a wider angle on people’s decision-making motives.

3

Chapter 1

The Effect of Effort in The Ultimatum Game:

A Cross-Cultural Analysis

Bassim Abdullah H. Allaheeb

With: Professor Luigi Mittone

and Professor Matteo Ploner

4

Abstract

The Ultimatum Game (UG) has been used as a constructive experimental tool for understanding human behavior in terms of decision making as well as certain other aspects, for instance,

rationality, and also issues related to fairness and punishment. This study examines the impact of effort on subjects’ behavior in the UG. More precisely, it analyzes not only the impact of

responders’ effort on proposers’ decisions but also the effect on responders’ decisions

themselves. For this purpose, two lab experiments were run in Italy, China, and Saudi Arabia. In the first experiment, responders had to exert a certain amount of effort to be able to claim an amount of the endowment while responders in the second experiment had an option to claim a proportion without exerting effort. The latter experiment was used as a control group for the first experiment. These experiments were used to observe the extent to which hierarchy or property rights have an impact on subjects’ behavior. This study found that when Italian responders exert effort, they significantly reject low offers more than when effort is not exerted. It also concluded that endowments in the Italian experiments were observed as property rights, while Saudis were influenced by hierarchy. On the other hand, the Chinese experiments revealed an approach towards endowments which lay mid-way between the approach in Italy and in Saudi Arabia. The study also found that when Italian responders exert effort, proposers significantly lower their offers more than when effort is not exerted. The latter behavior was not detected in the Chinese and Saudi experiments.

5

1.1. Introduction

In the Ultimatum Game, proposers are given the opportunity to decide a proportion of an endowment to be offered to responders, who have the power to accept or reject that offered proportion. If a responder rejects the offer both parties will receive nothing, whereas the endowment will be shared according to the proposed offer if the responder accepts (Guth et al, 1982). Research has documented various studies addressing various effects observed during the UG. For instance, the UG has been implemented in experimental studies from a range of

perspectives, and results have varied depending on the emotional, demographic, age, or gender variables on which each study focused. To mention a few, Kagel and Moser (1996) investigated fairness in the UG, Bornstein and Yaniv (1998) examined individual and group behaviors in the UG, Solnick and Schweitzer (1999) studied the impact of physical attractiveness and gender on UG decisions, Weber et al., (2004) explored timing and virtual observability, Oosterbeek et al., (2004) looked at cultural differences in UG studies. However, in this paper, this study only concentrates on utilizing UG as an effective tool for predicting human economic behavior regarding decision-making. More specifically, the first experiment group addresses the role of responders’ effort in increasing (or decreasing) the proposers’ offers, as well as its effect on responders’ decisions toward accepting or rejecting offers. By collecting data from three

different countries this study concluded whether the endowments were perceived as representing ownership rights or hierarchy in a cross-cultural perspective.

Most UG studies have concluded that responders tend to “take costly actions that express their concerns for fairness by rejecting positive offers” (Carr & Mellizo, 2013). This suggests that responders in behavioral economic games take fairness into account. Therefore, a proposer should strike a balance to ensure this fairness by raising the proportion of the amount offered to

6

the responder. Fairness should also be present in proposers’ offers when responders exert effort. Guth and Kocher (2014) provide another logic for rejections in that “responders do not only care about their own monetary payoff but compare their payoff with that of the proposer and become frustrated when their share is much lower”. In this study, effort is used as a tool for examining proposers’ and responders’ decisions.

Since there is a lack of research addressing the role of responders’ efforts in affecting their decisions, this paper aims not only to bridge a gap in the literature of behavioral economics, but also to study perceptions of endowments from the point of view of different cultures and discover whether culture plays a significant role in the UG or not. More precisely, to what extent are endowments perceived as property rights, or as a matter of hierarchy in the light of exerting effort?

The differences in perceiving the experiments’ endowments in three countries of three different cultures (Europe, Asia, and Middle East) were studied in this paper. The findings related to these countries, namely Italy, China and Saudi Arabia, can be generalized to other countries or at least be considered as a reliable source of understanding of the effect of cultural differences.

1.2. Literature Review

Since its introduction by Güth et al. (1982) the UG has been discussed extensively in literature in the field of economics, especially, in behavioral economics. The game rejections made by the responders to low offers has shed some light on ignored aspects of human behavior involved in decision making, fairness, punishment, cultural effects, self-control, and so on. For instance, responders’ acceptance of any offer is greatly influenced by their sense of not being taken advantage of, or as what the literature called fairness (Fehr et al., 1993; Burrows & Loomes,

7

1994; Forsythe et al., 1994; Kagel et al., 1996; Fahr & Irlenbusch, 2000; Camerer, 2003). To ensure that this sense of fairness persuades responders to accept the proposed offers, this study relies on responders’ efforts to clarify the effects of effort on proposers’ offers and whether effort plays a role in responders’ decisions or not. One noteworthy issue which arises from this

experimental study involve other related variables, more precisely, social value orientations (SVO), the main focus, being on the effort variable in the experiment group.

By introducing effort to the UG, the researchers in this study were able to correlate efforts to the offered proportions, in which responders exerted some amount of effort to be able to claim a certain proportion. However, subjects in the experiment adjusted their behavior according to the effort exerted thus showing how effort plays an important role in the UG decisions for both parties, proposers and responders. Since effort has not been extensively examined in the UG, it is useful to analyze how effort has been investigated in other similar games. Several previous studies have emphasized the impact of making considerable effort to accomplish bargaining tasks. Most of these studies have concentrated on the way endowments are obtained by responders, who are impacted by proposers’ spent effort. In one notable study, Burrows and Loomes (1994: p. 203) concluded that, in reciprocity games, “differentials in payoffs are deserved, and are, therefore, fair if they correlate positively with the amount of effort involved in obtaining them”. In the same vein, Fahr and Irlenbusch (2000) noticed that

responders in trust games were willing to accept offers when effort was exerted by proposers. This therefore suggests that effort, regardless of what form of effort is spent, plays a vital role in increasing or decreasing the likelihood of responders’ willingness to accept or reject an offer. Similarly, Gintis et al., (2003) found some evidence demonstrating that, in an employer-employee game, the more the employer-employee role-player provides effort (e), the more the employer

8

role-player will raise the wage (w). In other words, Gintis et al., (2003) noticed that effort in behavioral economic games is a key to affecting bargaining rates. Moreover, in the moonlighting game, Abbink, Irlenbusch, and Renner (2000) examined the effect of effort on enhancing a moonlighter (responder)’s acceptance of an offered endowment. They pointed out that effort contributes to increasing responders’ trust in proposers, and thereby reaching successful deals with them. A study by Fehr & Schmidt (2001) examined the effect of effort exerted by

responders in the Gift Exchange Game on the proposers’ decisions. Carr and Mellizo (2013) utilized the UG to identify some factors that impact responders’ acceptance or rejections to offers other than the distribution of endowments. They concluded that the source of endowments has an impact on the proportions of acceptance and/or rejections. In their study, they provided different endowments based on the proposers’ efforts. A paper by Bland et al (2017) mentioned that “In

real life, however, people’s propensity to engage in cooperative behavior depends on their effort and contribution; factors that are well known to affect perceptions of fairness.”. Carr and

Mellizo (2017) found that the probability of rejecting an endowment is lower by more than 44% when endowment is generated by responders’ effort than when endowment was exogenous. In this study, as mentioned earlier, the endowment was fixed, but the proportion a responder could claim was linked to some amount of effort. This procedure ensures a direct relation between the exerted effort, the proportion claimed, proposers’ offers, and responders’ final decisions.

This paper examines the effect of the responders’ claiming option and whether that claim was done with effort or without, and how it impacted on the participants’ decisions. Using this option has helped us to understand the different kinds of perceptions of endowments more fully and whether they are perceived as property rights or there is a hierarchy effect on participants’ decisions.

9

1.3. Research Gap

There has been little research addressing the role of responders’ effort in persuading proposers to modify their offers in the UG. Linking efforts to the claimed proportions gives us a more realistic interpretation of human behavior; whereas the addition of claiming options has enabled us to identify the motives of responders’ decisions and give clearer interpretations about endowments, regarding the extent to which they were perceived as property rights or as a matter of hierarchy among the different cultures. The central objective of this study is to observe the players’ perspectives and understand the real motives behind certain decisions. In this way it will contribute to the literature, thus bridging the research gap in this field. Therefore, the differences between three countries from Asia, Middle East, and Europe, were studied to be able to

generalize the findings and to have a wider consideration about different cultures on behavioral decisions that are related to fairness.

1.4. Target Population:

This study was run in three different countries, as mentioned earlier, for two main reasons. The first reason is that the target population represents some similarities and some differences regarding their backgrounds, financial systems, languages, and three other aspects as follows: Firstly, all the three countries are members of the G20 (Group of Twenty) which is an

international forum to discuss global economies by the top economy countries. Secondly, People from all the three countries belong to or affected by different religions, i.e. in Italy, which has the Vatican City within its borders, most people are influenced by Christianity, mainly the Catholic Church, while in China most of the population are either atheists or Buddhists, whereas people of Saudi Arabia are all considered Muslims except some of non-Saudi workers who are working in Saudi Arabia with a working-visas. The country also has the two holiest mosques of Islam,

10

Makkah and Madinah, and the direction of all Muslims around the world when they want to pray or perform Hajj, which is the fifth pillar of Islam. Thirdly, all people from the three countries speak different languages, namely, Italian, Chinese, and Arabic. Fourthly, Italy, China and Saudi Arabia are all known for being ancient nations. Although the country of Saudi Arabia got its current name in 1932 as the Kingdom of Saudi Arabia its roots go back up to 130,000 years ago (Armitage et al., 2011). Fifthly, each country represents different location and different financial system, which both have an impact on people’s lives. For instance, Italy represents a western country that is mainly run by capitalism, while China represents as an eastern country that is transforming in its financial systems, and Saudi Arabia as a Middle-Eastern country that is run by Islamic financial system. Sixthly, all the three countries have different dimensions in the Hofstede Model, more precisely, the Power Distance dimension (PDI). The Hofstede Model proposes six different dimensions, which represent the effect of culture on individuals’ value (Hofstede, 2011). The targeted population in this study is compared according the PDI

dimension, which is defined by the author as “the extent to which the less powerful members of

institutions and organisations within a country expect and accept that power is distributed unequally”. Table 1 summarizes the first reason behind the motives of choosing the three

countries: Table 1

Differences between the three chosen countries

Italy China Saudi Arabia

G20 Ranking 9th 3rd 17th

Religion Influence Christianity - Catholic Atheists (non-religious) Islam - Sunni

Language Italian Chinese Arabic

Heritage Ancient Ancient Ancient

Location &

Financial System Western - Capitalism Eastern - Transforming Middle Eastern - Islamic Financial

11

From the table above, it can be seen that each country represents a unique population that shapes people’s choices and may lead people to make decisions in different ways. For instance, the PDI in the Hofstede Model shows that Saudi Arabia’s score is 95, which is a very high score,

exemplifies how less powerful people are willing to accept unequally distribution of power. This score also tells us that people of Saudi Arabia are such of hierarchy oriented, centralization is accepted and those less power people are expected to accept what they are told to do. In contrary, Italy’s score represents unbalanced levels of PDI, that is, people from the northern part of Italy comparing to the southern part prefer to have decentralization and equally distribution of power more than those of the southern (Hofstede, 2011). Furthermore, Chinese people tend to accept inequalities amongst people. According to Hofstede’s interpretation, Chinese people are “influenced by formal authority and sanctions and are in general optimistic about people’s

capacity for leadership and initiative.”.

The second reason is that the target population characterizes different point of views about exerting effort. Firstly, according to Statista (2017), Chinese people ranked at the top of the average number of steps people took per day with 6,189 steps per day. Whereas the average Italians steps per day was 5,296. Both Chinese and Italians people spend higher amount of physical activity comparing to Saudi Arabian people whom average number of steps took per day was only 3,807. Secondly, according to WHO’s Insufficient Physical Activity Index (2018) China has a score of 14.1% as an insufficient physical activity amongst people who are older than 18 years old which puts Chinese people at the top in the list as active people, while Italy has a score of 41.4% which means that Italians are not as active as Chinese people but they are much better than Saudis, who are in bottom of the list with a score of 53.1%. Furthermore, there is a difference between the southern and northern Italians regarding productivity, in which, the

12

northern Italians productivity is higher than southern Italians (Giunta, Nifo & Scalera, 2012) and this study was conducted in the northern part. While China’s rapidly developing and changing systems are pushing towards exerting more effort to meet those changings it is reported that Chinese are working more than the legal standard hours per week (Zeng, et al. 2005). However, exerting effort in Saudi Arabia is seen as a not a favorable thing to do. A study by Al-Hazza (2007) concluded that the majority of Saudis were inactive regarding exerting physical efforts and that more than 72% of his sample did not engage in any physical activity that last only for 10 minutes.

To sum up, here are the two experiments:

1- UGT: treatment group in which responders must exert a certain effort to be able to claim a certain proportion of the endowment.

2- UGC: control group in which responders can claim a proportion without exerting effort.

1.5. Research Questions

This study’s main research question is: Do subjects from different cultures perceive endowments as property rights or as a matter of hierarchy?

In order to answer the primary research question, analyzing the two following secondary questions was needed:

1- Does the UG with effort plus the claiming option have a larger influence on the number of rejected offers in the UG compared to just the claiming option? More precisely, would low offers that are usually rejected in the UG with effort, including the claiming option, be accepted in the situation with only the claiming option (without effort)?

13

2- By knowing that responders have exerted effort to be able to claim certain proportions, would proposers offer a lower proportion to responders than they would without responders’ efforts? (Since responders have exerted efforts, they might not want to lose that effort so they might accept lower offers).

1.6. Methodology

This study was conducted experimentally in three different countries. In Europe, the experiments were run in Italy, at the CEEL lab at the Universita' di Trento. In Asia, the experiments were run in China, at the FEEL lab at Xiamen University and in the Middle East, the experiments were run in Saudi Arabia, at Majmaah University. The study includes two different experiments. Firstly, the UGT (with the effort component plus claiming option) was used as an experiment group. Secondly, the UGC (with only the claiming option and without exerting effort) was used as a control group. A binary regression analysis as well as an

independent-samples t-test analysis were used to detect the significance of each variable in both experiments.

1.7. Hypotheses:

Hypothesis 1: Claims with exerted effort by responders (UGT) significantly and negatively influence proposers’ decisions about the offered size of the proportion of the endowment in the UG.

The first hypothesis states that having both effort and the claiming option significantly and negatively influence proposers’ offers compared to having just the claiming option. To test this hypothesis, an independent-samples t-test analysis was drawn to detect the offers mean

14

Hypothesis 2: In the UG with responders’ effort plus claiming option (UGT), in societies with stronger hierarchical structure – i.e., lower impact of property rights – low offers are more likely to be accepted.

The second hypothesis states that there is a significant and positive relationship between

accepting low offers and cultures that are based on hierarchy in the UGT. To test this hypothesis, a binary logistic regression was conducted using the data from the two experiments (UGT and UGC). The dependent variable was the acceptance of offers and the independent variables were the proposed offer, claim-offer difference, and gender.

1.8. Experimental Design

1.8.1 Data Collection

Data was collected by recruiting 308 subjects randomly from the three countries (Italy: 140; China: 52; Saudi Arabia: 116) as shown in Table 2 below:

Table 2

Number of participants from each country

UGT UGC UGT Total UGC Total Total

Male Female Male Female

Italy 49 39 22 30 88 52 140

China 17 19 7 9 36 16 52

Saudi Arabia 48 12 49 7 60 56 116

Total 114 70 78 46 184 124 308

These subjects were mostly university students from different majoring fields, thus ensuring to have different backgrounds in our data. A show-up fee of 3 Euros was given to each Italian participant as encouragement to participate in the study, and an endowment of 10 Euros was given for bargaining. Subjects from China and Saudi Arabia were given the same value based on

15

the Big Mac Index. The total budget of the six experiments (3 treatments and 3 control groups) was approximately 2,470.00 Euros1_.

1.8.2 Laboratory Environment

Prior to conducting the study, the participants in all experiments had to read detailed instructions about the study rules including the 10 Euros endowment (or its equivalent in other currencies based on countries’ purchase power of Big Mac Index), how this amount is bargained, how efforts are made, when deals are made, and when both players lose the entire endowment. The instructions were read before the start of the game and were shown on the computers that the subjects used. For instance, each participant read that if the responder accepted the offer, the endowment would be shared out according to the agreed proportion. However, if the responder rejected the offer, both bargainers would lose the whole endowment. To make sure there were no other effects, all offers were made under conditions of anonymity. In other words, proposers did not know the identity of the responders they were matched with and vice versa. To implement this method effectively, proposers and responders did not know about their roles until they read the instructions on their computer screens when their roles popped up. Moreover, the

participants’ performance was monitored by providing them with feedback or help whenever needed. Furthermore, the experiments were originally written in English then translated into 3 different languages: Italian, Mandarin, and Arabic then all those three translated experiments were translated back into English by other translators to ensure that there was no ambiguity in the words being used in the experiments. During the setup of laboratories, there was no issues in both the Italian and Chinese experiments, but in Saudi Arabia there was a problem having males

1 _{This does not include other expenses such as travel expenses, printing documents, assistance} fees, and other lab expenses.

16

and females to present together in the same laboratory. The education system in Saudi Arabia is mainly forcibly segregated by gender from the elementary school to graduate studies

(Alsuwaida, 2016) with some few exceptions like King Abdullah University of Science and Technology (KAUST) and in some majors in other universities like medicine. With the current reforming in the country, some graduate programs have started to eliminate such segregation. In this study, it was impossible to recruit participants in the same laboratory, instead, males

participants were placed in a lab while female participants were placed in another one in the females section. With this type of restriction it was not easy to recruit as many females as males and that caused an unbalanced number of participants among the Saudi participants. All games were developed and implemented by using the Otree platform, which is an open-source software platform used for laboratories, online, and fieldwork (Chen, Schonger & Wickens, 2016). 1.8.3 Procedures

Subjects in all the three experiments were told about an endowment of 10 Euros (or its equivalent in other currencies based on a country’s purchasing power on the Big Mac Index) to be negotiated between two types of subjects, proposers and responders. Yet, both proposers and responders were assigned randomly, and each proposer was automatically paired with a

responder by the computer (generated by default). This signifies that, once the proposer sends his/her offer, the computer randomly delivers the offer to a responder. More importantly, the computer makes an explicit linkage of each pair so that tracking their bargaining processes was possible. As a clue, responders were told how much proposers have proposed and also how much effort responders have made and how much their claim was shown to proposers. After that, if a responder rejected an offer, the proposer was notified about this rejection on his/her computer screen. In the same manner, if the responder accepted the offer, the proposer was also notified

17

that the deal was made successfully. Then, all subjects’ decisions were collected and analyzed in order to identify to what extent responders’ efforts influenced the two parties’ decisions. Finally, the findings of the first experiment (UGT) were compared with a separate control group (UGC) where no effort was involved except the ability to claim.

1.8.4 Offer Making

In the experiments where responders exert effort (UGT), responders were given five minutes to exert effort by performing a slider-task on the Otree platform. The slider-task pops up on the screen and responders’ effort is made when responding. In other words, the slider-task is considered as an effort that responders must make in order to claim a certain proportion.

Responders cannot select the next proportion level until they make a certain effort. To make this clearer, Table 3 illustrates how the proportions and efforts are laid out in the UGT. Thus,

responders consider these efforts carefully prior to making their decisions. Table 3

Efforts made by Responders (B), and amounts can be claimed from Proposers (A) Sequence Efforts (Sliders) Claim (B)

1 50+ 1, 2, 3, 4, 5, or 6 2 40 －49 1, 2, 3, 4, or 5 3 30－ 39 1, 2, 3, or 4 4 20－29 1, 2, or 3 5 10－19 1, or 2 6 0 －9 1

According to Camerer (2003) an offer of 50% is usually accepted in standard UG studies (which reflects offer #2 in Table 3), and any proportion given to a responder greater than 50% is

considered rare (#1 in Table 3) as well as offers of 10% or lower (which mirror offer #6 in Table 3). Similarly, another experiment was performed, yet without including the effort variable, which is UGC. This experiment was used as a control to the first experiment (UG with effort).

18

1.9. Results

Three experiments and three control groups were run in Italy, China, and Saudi Arabia. Italian sessions were done in May, June, and October of 2016, and Chinese sessions were done in July 2016, whereas the Saudi sessions were run in the April and May of 2017. During the sessions, we were assure that they do not overlap with external effects like religious days, exams, or other factors, also, we had not faced any IT problems in any of the sessions. To better show the results of those countries, graph 1 contains each experiment from each country with proposers’ offers in Y columns and responders’ claims in X rows, while red triangles are the rejected offers. Sections 1.9.1. and 1.9.2. are experiments results in details.

19 UGT UGC P ropos er s’ O ff er s C hi na It al y S audi A ra bi a

UGT Responders’ Claims UGC Responders’ Claims Graph 1: Proposers’ offers and responders’ claims in the UGT and UGC (red triangles

20

1.9.1. The UG with responders’ claims and efforts (UGT) experiment:

In this game, responders must exert effort to be able to claim a proportion. If a responder decided not to exert any effort, s/he can only claim a proportion of 10%. However, if a responder wanted to claim a higher proportion, s/he would have to exert more effort. For instance, if a responder wanted to claim a proportion of 50% s/he would have to exert an effort of at least 40 slider- tasks. Unsurprisingly, none of the 4 or 5 offers were rejected in all of the experiments carried out in the three countries except in the control group of Saudi Arabia in which some users rejected fair or even generous offers. This latter kind of behavior might be caused by the subjects’ own beliefs and the influence of religion which these subjects had already effectively rejected by participating in these experiments, fearing that this type of experiments might violate their religious teachings. Here are the main results of the UGT:

In Italy, eight of the offers were rejected. All of them were three or lower, and all of the rejected offers were equal to or lower than responders’ claims. While only two offers were rejected in Saudi Arabia , all of them were also three or lower. However, only one offer was rejected in China and that offer was as low as two. By examining the Italian results, it was found that exerting effort plays a significant role in responders’ decisions. This result is shown by

Table 4

UG with effort and claim option (UGT)

Country N Gender Age Task Avg. Proposers’ Avg. Offers

Responders’ Avg. Claim

% of rejected offers Male Female Min Max

Italy 88 49 39 20 28 37.45 3.57 3.95 18%

China 36 17 19 16 25 40.33 3.72 4.39 5.55%

21

comparing responders who exerted effort (as in UGT) with responders who did not (as in UGC). On the one hand, in the Italian UGT there was a positive significant relationship between the proposed offer and its acceptance (p=.039), and a significant negative relationship between the difference of claims to offers and accepting offers (p=.074) which means that when the

difference value between the offer and the claim was negative the more probability the offer would be rejected. In addition, a positive significant relationship was found between being a male and accepting offers (p=.078) which indicates that women are not as likely to neglect efforts. In other words, a female responder in the UG who exerts a certain amount of effort would be highly expected to reject any offer below her claim unlike males who are more likely to accept lower offers. On the other hand, the Chinese and Saudi experiments revealed no significant relations between the independent and dependent variables when using the binary logistic regression. However, when using a descriptive analysis, the Saudi experiments revealed unexpected behavior in the UGT where no effort was exerted to claim a specific proportion. This kind of behavior can be interpreted in four possible ways: (1) it may be the result of laziness, since Saudi Arabia ranks as being in the top three laziest countries in the world according to the World Health Organization (WHO) publication (Guthold, Stevens, Riley & Bull, 2018). (2) It may be due to the subjects’ religious beliefs which might affect their decision choices in the game, i.e. the Islamic religion can effectively impact the players’ behavior as it prohibits the use of certain activities such as gambling, uncertainty, ambiguity in trading, and earning interest on savings2_{. However, the first reason seems more feasible seeing the considerable number of}

2 _{In Chapter 3 of the Thesis, a detailed investigation about the impact of Islam religion on} experimental studies about decision making will be provided.

22

proverbs and sayings present in the Saudi culture that subtly encourage inertia3_{. (3) University} students in Saudi Arabia receive monthly salaries from the universities ($266 per month) and they do not pay any tuition fees. The given show-up fees in the experiments as well as the allocations might not considered as a good incentive to them to exert effort or to take the experiment seriously. Cameron (1999) and Andersen et al (2011) concluded that the amount of the stakes in the UG play a major role in players’ decisions, which makes responders to be more willing to accept offers if stakes were high offer. (4) The unbalanced recruited gender of

participants might cause this type of behavior. (5) The lack of experience and knowledge of experimental studies among students in Saudi Arabia might also be the causal of having this kind of result, which might led them to take the experiment as a joke and not seriously. A further study about number (4) and (5) is needed to expand our knowledge about these kind of effects on participants’ decision-making process.

From graph 1 above, the Italian and Chinese responders’ claims with the amount of effort they expended is correlated. The greater their effort, the greater their claim. The red triangles are the rejected offers. However, the Saudi Arabian graph shows the claim at its lowest (1 point) which means Saudis did not exert any effort to claim any point higher than 1. This suggests that Saudis

3 _{For instance, “What is not being done today will be done tomorrow” this proverb is used as a} pretext for relaxation; “One bird in the hand is better than ten birds in the tree” is said when someone wants to attempt to gain profit or reward through hard work but others try to change his/her mind; “Relax your angels” means not to bother yourself in any undue effort, just relax; the proverb “What costs you less is blessed” is used to encourage people to spend less money or expend less energy on any aspect of life; “Relaxation is half of our daily bread ” is said to encourage someone to relax when exerting effort to earn more money; “If you ran like monsters you wouldn’t gain more than has been decided for you” this proverb means that no matter how hard you try you will not achieve more than what you are achieving now; “What is burning your rice?” is said to someone in a hurry in an attempt to calm them down a little; There are many more other proverbs that are well-known in the Saudi culture that promote relaxation and warn against working too hard.

23

reacted differently from Italians and Chinese participants. Effort represents a tool to enable us to identify whether participants consider endowments as a property right, where responders exert effort to claim their shares, or as a hierarchy, where responders may not become involved due to the effect of the power distance. For instance, high scores in Hofstede’s model mean a great distance in power and inequality between a manager and an employee, an older and a younger person, and so on. Hierarchy in those countries whose scores are high in the model is clearly present. Since the second player has the power to accept or reject the offer of the first player the impact of this power distance is detected by a failure to exert effort as shown in the UGT. In addition, by comparing the UGT with the UGC, it was found out that the Italian responders in the UGT had the highest rejections rate among the three countries followed by the Chinese responders and lastly the Saudi responders. Whereas in the UGC the opposite was found, in which, the highest rejection rate was given by the Saudi responders followed by the Chinese responders and lastly by the Italians. The only difference between the UGT and the UGC was the effort put in. This tells us that the Saudis were influenced by hierarchy and that the Italians where influenced by property rights. Furthermore, by analyzing the Chinese responders, it was found that they are more likely to be influenced by hierarchy than by property rights since the rejection rate was low in the UGT and high in the UGC.

1.9.2. The UG with claims but without effort (UGC) experiment:

Table 5

The UG with claim option (UGC)

Country N Gender Age Avg. Offers Avg. Claim % rejected offers

Male Female Min Max

Italy 52 22 30 18 33 4.46 4.85 3.85%

China 16 7 9 18 24 4.5 5 12.5%

Saudi Arabia

24

In this control group, this study shows that the percentages of the rejected offers are inversed compared to the UGT. In the UGT the Italian rejection rate was the highest, while in the UGC it was the lowest. The same inverse results were seen with the Chinese and Saudi results. Both countries’ rejection rates were low in the UGT but high in the UGC. This suggests that the perception of the endowment was affected by culture. This study assume that Italians were more oriented towards receiving property rights in perceiving endowments while Chinese and Saudi Arabian subjects were more likely to be hierarchy-oriented. This result corresponds to the Power Distance (PDI) dimension in Hofstede’s model (Hofstede, 2011) in which the Italian score of the PDI is 50, the Chinese is 80, and the Saudi is 95. This dimension of the model is interpreted as scoring higher in a country where the distance between the decision maker and the follower is greater. It is defined by Hofstede as “the extent to which the less powerful members of

institutions and organizations within a country expect and accept that power is distributed unequally”. Therefore, individuals in Saudi Arabia are not expected to reject any hierarchy order which is reflected in this study as the rejection rate in the UGT compared to the UGC. When responders were obliged to exert effort in the UGT, it was implied that these players were beta subjects and not decision makers even though they had the ability to accept or reject offers. On the contrary, when no effort was necessary in the UGC, responders behaved as alpha subjects. By alpha subjects we mean those who have hierarchical power, and beta are less powerful subjects. Rejecting offers in the UGC reflects how responders in Saudi Arabia and China were aggressive even when the offers were generous. This type of behavior is absent in the Italian UGC but present in the UGT because, as mentioned earlier ,Italy has a low score in the PDI compared to China and Saudi Arabia.

25

A logistic regression was run to check the effect of effort on responders’ behavior. By including claims, and gender variables it emerged that Italian responders significantly reject offers that are lower than their claims when they were associated with effort (p= .039) and the standard error is 0.38. Furthermore, women tend to reject offers significantly more than men when claims are higher than proposals (p=.078) and the standard error is 0.96. In addition, by correlating Italian claims to offers, results showed that an offer of 3 was not always accepted. Interestingly, the Italian UGT showed that when responders exerted a certain amount of effort in an attempt to claim a certain amount of endowment, 31% of the responders chose not to claim the maximum available option. In the same vein, 16.67% of the Chinese responders

demonstrated the same behavior. For instance, if a responder exerted an effort of 43, which enabled him/her to claim 50% of the proportion, s/he claimed 40%. This behavior was not detected in the UGT of Saudi Arabia because all participants did not exert such efforts. To sum up the Italian results from both the UGT and the UGC, this study concludes that there was a significant effect of effort when responders exert effort and claiming was considerable (see the UGT analysis in Table 5 below). This effect was not detected when responders only claimed certain proportions without exerting effort (as in the UGC). These results suggest that Italians perceive endowments as a property right rather than a hierarchy. In contrast, there was no significant effect of effort with claiming on responders’ decisions in Saudi Arabia and China as well as no significant relationship between claiming and responders’ decisions in the

experiments with no effort exerted. Since the sample size of the UGC in China is relatively small, a Fisher’s Exact Test was run and it did not reveal any effect on these variables too. This result strengthens the study’s hypothesis regarding the perception of endowments as a matter of

26

hierarchy in cultures that are mainly influenced by hierarchies. Hence, the null hypothesis 2 is rejected.

Table 6

Binary Logistic Regression for the UGT

Italy: B S.E. Wald df Sig. Exp(B)

Offer 0.84 0.407 4.253 1 0.039 2.317 Claim-Offer Diff. -1.024 0.574 3.181 1 0.074 0.359 Gender 1.831 1.04 3.099 1 0.078 6.243 Constant -0.989 1.16 0.727 1 0.394 0.372 China: Offer 92.275 7989.136 0 1 0.991 1.19E+40 Claim-Offer Diff. -30.624 2991.901 0 1 0.992 0 Gender -125.641 12931.28 0 1 0.992 0 Constant -74.153 11747.4 0 1 0.995 0 Saudi Arabia: Offer 0.392 0.505 0.603 1 0.437 1.48 Gender 1.136 1.585 0.513 1 0.474 3.113 Constant 0.637 1.561 0.166 1 0.683 1.89 Table 7

Binary Logistic Regression for the UGC

Italy: B S.E. Wald df Sig. Exp(B)

Offer 33.229 5802.252 0 1 0.995 2.6999E+14 Claim-Offer Diff. 3.111 17572.655 0 1 1 22.434 Gender 70.958 20745.996 0 1 0.997 0 Constant -48.676 13692.48 0 1 0.997 0 Saudi Arabia: Offer -0.008 0.432 0 1 0.985 0.992 Claim-Offer Diff. -0.024 0.297 0.007 1 0.936 0.976 Gender -0.362 1.24 0.085 1 0.77 0.697 Constant 1.451 2.089 0.482 1 0.487 4.266

To detect the impact of responders’ effort on proposers’ decisions, the offers mean differences between the UGT and UGC was conducted by using the independent-sample t-test. This study

27

found a significant and negative difference between the Italian UGT and UGC offers (P=.006)4_. Which means that offers were significantly lower when responders exerted effort (UGT) than when they did not exert effort (UGT). Italian proposers lowered their offers in the UGT because they knew that responders had had exerted adequate amount of effort to claim certain amount of the proportion, and that rejecting an offer would cause unpleasant feeling more than if the claim was done without effort, hence, offers in the Italian UGT were significantly lower than in the UGC. In other words, since responders in the Italian experiment did not exert effort in the UGC they were offered higher amount of proportions than in the UGT because proposers were more worried about rejecting their offers in the UGC than in the UGT. On the contrary, the offer mean differences in the Chinese and Saudi experiments revealed no significant difference between the two experiments (P=.124, P=.775) although the Chinese and Saudi offers averages had the same attitude as the Italian, which show that the averages of the UGT offers were lower than the averages of the UGC. Therefore, the null hypothesis 1 is rejected for the Italian society but it is not rejected for the Chinese and Saudi experiments.

1.10. Social Value Orientation (SVO) :

A Social Value Orientation (SVO) task was given at the end of the experiments. SVO is a well-known tool that measures the magnitude of the concern a person has for others. The SVO scale ranks as follows: Competitive, Individualistic, Prosocial, and Altruist. In short, competitive is the minimization of others’ payoffs regardless of the self-payoff. On the contrary, Altruist represents the maximization of others’ payoffs regardless of the self-payoff. However, Individualistic represents maximizing self-payoffs with respect to others’ payoffs. Lastly, Prosocial represents maximizing others’ payoffs with respect to self-payoff. From collected data, it was found that

28

Italian participants were the highest in the ranking of the Prosocial measure, then Chinese, then Saudis. Conversely, Saudis were the highest in the Individualistic ranking, which supports why Saudis rejected exerting effort when other participants, whom also expected to be individualistic, did not have to exert effort, hence, Saudis preferred to not exert effort. followed by the Chinese, then the Italians. There was no Altruists or Competitive participants in the experiments.

Correlating the results of the SVO with the rejections in the UGT experiment shows that the number of Italian Individualistic and Prosocial responders both amounted to 4 (4 Individualistic, and 4 Prosocial). On the other hand, Chinese responders who rejected the offer were Prosocial. Finally, Saudi responders who rejected offers were all Individualistic. To sum up, the percentage of those who rejected the offers was 54.5% Individualistic and 45.5% Prosocial. Here is a table showing the SVO patterns in percentages:

Table 8

SVO frequencies

Categories Italy % China % Saudi Arabia %

Altruist 0 0% 0 0% 0 0%

Prosocial 25 56.82% 9 50% 7 23.33% Individualistic 19 43.18% 9 50% 23 76.67%

Competitive 0 0% 0 0% 0 0%

Total 44 18 30 100%

29

1.11. Further Investigation

In this study, the impact of responders’ effort on both the proposers and responders behavior was tested. A further analysis will be provided in the second and third chapters of this thesis. The second chapter will discuss the impact of proposers’ effort on the two players, the proposer and the responder. Whereas the third chapter investigates the impact of people’s attitudes in Saudi Arabia as an Islamic society toward the experimental studies.

1.12. Recommendations for future studies

Based on the results from the Saudi Arabian experiments, conducting a comparative

experimental study to detect the differences between experiments that violate religion beliefs and experiments that do not violate religion would reveal more understanding on the impact of using behavioral economics terms in the experimental studies that have been ignored up to our

knowledge. Researchers are encouraged also to examine different parts of a country in a study to shed light on the differences between participants from different cities in a country on a certain topic regarding exerting effort in the UG. For example, conducting a study in the south and north of Italy may reveal differences towards the perception of the endowment as well as an industrial city in China, like Beijing, and a rural city may also reveal different perceptions about the

endowments. Another recommendation is to design the UG including the added effort variable in strategic choices, in which, proposers and responders can be asked about their decisions on each available choice before showing the other party’s decisions.

1.13. Limitations

Since laboratory studies cost money, it was not feasible to expand this study to involve other countries or to include other cities within the chosen countries. Another limitation is that this

30

study was targeting mainly university students and did not include other types of participants like employees whom might have different behavior towards exerting effort in the UG.

1.14. Conclusions

In this study, exerting effort by responders was introduced to the Ultimatum Game as a tool to detect the perception of endowments in the game. The experiments of this study were run in three different countries, Italy, China, and Saudi Arabia. The treatment group in the Italian experiment revealed significant relationship between the amount of the offered proportion by proposers and the claims of responders when effort was exerted. On the other hand, the results of the Chinese and Saudi Arabian experiments did not reveal any relation between the amount of offers and the claims by responders in both the treatment and control groups. Hence, the first null hypothesis is rejected for the Italian society but it is not rejected it in the Chinese and Saudi Arabian societies.

In addition, the Italian experiment group shows a significant relationship between the

responders’ acceptance to the proposers’ offers and responders’ claims that included efforts but did not reveal any relation in the control group where responders did not exert effort.

Nonetheless, both the Chinese and Saudi experiments did not reveal any relation between offers and acceptances, therefore, the second null hypothesis is rejected for the Italian society but it is not rejected it in the Chinese and Saudi Arabian societies.

The Saudi Arabian results in both experiments suggest that participants were not in favor of exerting effort. This kind of behavior tells us that Saudis were influenced by a certain motive which prevents them from exerting effort. That motive could be a person’s belief of violating religion or due to laziness or they perceived exerting effort as an extra input in which responders

31

felt unfairly treated. A further investigation about the Saudi Arabian society will be discussed in the third chapter to know more about the impact of religion on experimental studies.

Including effort to the Ultimatum Game has enabled us, in this study, to know more about the differences between cultures regarding how people perceive endowments. Societies with high hierarchy structures, like China and Saudi Arabia, behaved differently than property rights societies, like Italy. In which, the latter perceived endowments as property rights.

33

Chapter 2

Proposers’ Effort Impact in the Ultimatum Game:

A Cross-Cultural Analysis

Bassim Abdullah H. Allaheeb

With:

Professor Luigi Mittone

and Professor Matteo Ploner

34

Abstract

In the field of behavioral economics endowments have been used in experiments as a bargaining tool between participants. The main benefit of using endowments is the ability to identify how players are behaving within certain circumstances when they are negotiating to reach an agreement in a game. Real and hypothetical endowments have already been addressed in the literature so have not been a focus of this piece of research. This study has employed effort in the ultimatum game (UG) in three different countries to analyze the differences between proposers’ decisions when they exert some amount of effort optionally, and when it is a mandatory. It was found that the proposals were of lower value when proposers exert effort rather than when they did not. This behavior was interpreted as a way of guaranteeing that the responders would be convinced to accept the proposers’ offers. Furthermore, it was found that culturally driven phenomenon impact on the proposers’ behavior. Three experiment groups as well as three control groups were run in three different countries: Italy, China, and Saudi Arabia in order to detect the differences between the different cultures.

Keywords: experimental economics; Ultimatum Game; effort; culture; endowment; hierarchy; property rights.

35

2.1. Introduction

In the experimental economics it is broadly acknowledged that participants are hired to bargain for a certain reward, which varies according to the design of the experiment. The endowment has been discussed in various ways. For instance, Kahneman & Tversky (1979) provided an explanation of the endowment effect, where they indicated that both reference dependency and loss aversion are the two main features that account for the endowment effect. A decision maker in a transaction values an item according to a reference point s/he has in mind and each party, whether a buyer or a seller, perceive the item as a gain or loss. In this situation, loss aversion means the psychological impact of losing compared to gaining is greater even when the amount lost or gained is the same. Therefore, losing a $1 has a greater psychological impact on decision makers than gaining a $1. This phenomenon is further explained in the literature where it is argued that the greater impact of loss aversion is due to the fact that an owned item is of more value to the owner because it was chosen by them (Brehm, 1956). Further clarification is given by Morewedge, Shu, Gilbert, and Wilson (2009) who link the endowment effect to the ownership of an item rather than to loss aversion. They suggest that bargaining in the buying and selling procedure is correlated to ownership of the item being sold or bought which gives rise to the endowment effect. In their experiment, the endowment effect was not detected when the sellers sold unowned items and buyers purchased owned items. For example, if a buyer already owned an item s/he was ready to pay as much as sellers demanded for the same item. In the same vein, Kogler, C., Kühberger, A., & Gilhofer, R. (2013) found a significant impact of owning a lottery on the endowment effect although they suggest that the endowment effects in lottery settings is still a controversial topic in which explaining the real cause of the endowment effect in lottery settings might be due to anticipated regret rather than loss aversion. To achieve a more

36

in-depth evaluation, endowments can be analyzed in terms of whether they are real or

hypothetical. In an experiment by Kühberger, Schulte-Mecklenbeck, & Perner, (2002) the effect was identified even though the endowments were hypothetical, giving the same results that were presented Kogler, C., Kühberger, A., & Gilhofer, R’s paper (2013). In a contrary experiment, Loewenstein and Adler (1995) failed to identify a similar effect when the bargained endowment was a hypothetically owned mug, which people had failed to predict the accurate value of. Van Boven et al. (2003) contended that an intrapersonal empathy gap arises due to biased predictions from both parties, sellers and buyers. The prediction that a person is a buyer and the other person is a seller, or the opposite, creates this gap and as a consequence both parties underestimate the role of the endowment effect on each of them.

In the literature of behavioral economics, the ultimatum game (UG) has been used as a robust way to identify some decision-making effects among participants. The game was firstly introduced by Güth, Schmittberger & Schwarze (1982). The game is simple and reveals some aspects of human behavior that have been ignored. For instance, by employing the UG

researchers were able to observe the motives for rejecting positive numbers and agreeing to receive nothing, which in turn was interpreted as a way of expressing anger or punishing the unfair proposers. The game works as follows: a matching of two players who are referred to as a proposer and a responder. Then, an endowment is given to be split by the proposer who can decide the proportion for each player (self and others). For instance, the proposer can decide the proportion of the endowment to offer to the responder. The deal is considered successful only if the responder accepts the offer, otherwise, it is an unsuccessful deal. In classic economics, any positive number greater than zero is better than zero. Nonetheless, in the UG experiments it has been shown that not always do responders accept positive numbers especially when the offered

37

proportion is low, like 10% or 20% of the endowment. Actually, it has seldom occurred that responders accept very low offers. This kind of behavior has been seen in many other types of games in the field of behavioral economics. Research has investigated the motives of such behavior as mentioned earlier. This study contributes to the literature on understanding the perception of an endowment itself by using effort. How do people from different countries perceive an endowment? Do people from Italy perceive an endowment the same way as Chinese or Saudi people do? Two experiment groups were run to investigate the effect of adding efforts to the ultimatum game. Firstly, proposers in the first experiment (hereafter this text will be abbreviated as UGT) had a choice to either exert effort (Choice1) or alternatively simply choose not to exert effort (Choice2) to offer a certain amount of endowment. The outcomes of the second group (UGT-Choice2) were compared to the second experiment (hereafter this text will be abbreviated as UGC) in which proposers had to exert effort to be able to offer a certain amount of the given endowment to their matched responders. By employing this procedure, this study was able to compare proposers’ behavior when exerting effort was optional, and when it was mandatory, and when there was not effort involved to discover differences between the three groups in the three countries.

The abbreviations of the three groups are:

- UGT-Choice1: proposers’ effort is optional. - UGT-Choice2: proposers do not exert effort. - UGC: proposers’ effort is mandatory.

2.2. Existing literature

Exerting effort in the field of experimental economics have been used widely for various reasons. To mention some, Bull, Schotter & Weigelt (1987) used effort in a tournament

38

experiment in which subjects used how much to exert effort in order to compete between each other. van Dijk, Sonnemans & van Winden, (2001) used the effort in incentive systems for individuals and teams. Burrows, P. and Loomes, G. (1994) used the effort to test the impact of fairness on certain types of bargaining environments. Gneezy, Niederle & Rustichini (2003) implemented effort in order to detect the differences of performance between men and women. A paper by Bland et al (2017) mentioned that “In real life, however, people’s propensity to engage

in cooperative behavior depends on their effort and contribution; factors that are well known to affect perceptions of fairness.”. Sutter and Weck-Hannemann (2003) investigated the impact of

taxation on performance which was a type of exerting effort. Exerting effort in laboratories has a great advantage, that is, it gives greater validity of experiments to mimic the real-life effort (Gill & Prowse, 2011). A paper by Ruffle (1998) employed the effort in both the dictator and the ultimatum games. The effort in the latter game was used as a tool to know how much effort a responder had exerted to enable the proposer to receive a greater endowment in case the responder exerted an adequate amount of effort, and vice versa. Furthermore, MD Carr, P Mellizo (2013) included the effort in the ultimatum game to allow the responder to produce the endowment. Also, in the ultimatum game, García-Gallego, Georgantzís & Jaramillo-Gutiérrez (2008) used real effort in an employer-employee setting where exerting effort was exerted by the responders. Franco-Watkins, Edwards & Acuff (2011) applied effort to the ultimatum game, where earning money was via exerting effort rather than receiving money as a windfall. Effort was used also in other experimental games such as moonlighting game (Abbink, Irlenbusch & Renner, 2000). And is used in the trust game (Vilares, Dam & Kording, 2011). And in general, Dutcher, Salmon & Saral (2015) compared three types of efforts, real, stylized and trivial, in the public goods game, they found that all the three types revealed identical outcomes. The main use